Hey everyone! I'm currently working as a DevOps/SRE engineer and I want to help my friend who's been in AI/ML but is having trouble moving into better roles. Many job listings are now requiring skills like Kubernetes, CI/CD pipelines, Containers (like Docker or Podman), Infrastructure-as-Code (Ansible, Terraform), along with some knowledge of Linux and networking. My friend is really good with Python and various ML frameworks, but when it comes to hands-on experience with infrastructure, automation, and deployment, he's quite lost. I'm thinking of organizing a series of hands-on sessions, maybe one or two hours each week for a few months, covering real-world DevOps tasks. Here's my rough outline:
1. Basics of Linux & networking tools (like SSH, systemd, DNS)
2. Understanding digital certificates (OpenSSL, TLS, HTTPS)
3. Working with Containers (Dockerfiles, Podman, images, volumes)
4. CI/CD with GitLab or GitHub Actions (setting up test, build, deploy pipelines)
5. Learning Infrastructure as Code with Ansible and Terraform (enough to get him started)
6. Getting into Kubernetes (local setups with kind/minikube, basic manifests, and Helm)
7. Secrets management (using Vault, sealed-secrets)
8. Basics of monitoring and logging (Prometheus, Grafana, Loki)
I'd love to hear your suggestions:
- What else should I add or remove from this plan?
- Any beginner-friendly projects to tie everything together?
- How can I ensure he doesn't get overwhelmed while still focusing on what really matters?
- Any free resources or hands-on labs that you'd recommend?
Thanks a lot for your help! I'm really eager to set him up for success!
4 Answers
While your roadmap is impressive, keep in mind that only smaller companies expect knowledge of all these tools. Larger organisations usually have specialised teams for infrastructure and networking. Developers often won’t need to manage their own production setups. Just a thought!
Your plan sounds solid! Check out roadmap.sh/devops and devops-daily.com/roadmap—both are great for understanding DevOps step by step. For hands-on experience, spinning up servers with Docker or using managed Kubernetes on DigitalOcean is a fantastic way to dive in. They're affordable and perfect for beginners! Plus, DigitalOcean even offers GPU servers if he wants to integrate DevOps with his ML work, keeping it practical without too much setup hassle.
Your roadmap is on fire! 🔥 I suggest including a 'capstone' mini-project after every 2-3 topics (like deploying an ML model with Docker, then setting up CI/CD, and finally moving to Kubernetes). It keeps things real and builds confidence. For labs, check out KodeKloud, Play with Kubernetes, and GitHub’s devops-examples repo. Just keep your sessions focused—real wins are better than info overload!
The roadmap.sh/devops might give a good general view, but for a free hands-on approach, I'd recommend the Dynamic DevOps Roadmap found at devopsroadmap.io. It's a really pragmatic roadmap covering all aspects of DevOps without just listing tools.
Related Questions
Can't Load PhpMyadmin On After Server Update
Redirect www to non-www in Apache Conf
How To Check If Your SSL Cert Is SHA 1
Windows TrackPad Gestures